Browse and query web pages as a Unix-like filesystem with cached content.
npx skills add https://github.com/openprose/prose --skill websh
Please help me install this Agent Skill. Skill: websh Source: https://github.com/openprose/prose/tree/main/skills/websh Command: npx skills add https://github.com/openprose/prose --skill websh
Web browsing and data extraction are often manual, slow, and non-reproducible. The websh Skill solves this by turning the web into a Unix-like shell where URLs are navigable paths and content is queryable via commands.
Use Case: Build a reproducible workflow to fetch data from multiple sites, extract structured data, and save to local cache or export.
Start by launching websh, then run: "cd https://news.ycombinator.com" to explore Hacker News, then "ls" to see links, "grep 'AI'" to filter, "follow 1" to click a link, and "stat" to view page metadata.
High-intent search queries and answers about installing and using this skill.
Automate web data extraction by navigating to URLs with cd, caching HTML content, and progressively extracting it into markdown for fast grep searches. You can build reproducible workflows to fetch data from multiple sites and save it to a local cache.
Yes, you can browse live websites using a CLI interface by treating URLs as navigable paths. Commands like cd allow asynchronous background content fetching, while ls displays links and follow clicks them interactively.
Search and filter cached web content locally by using grep and grep-like commands on progressively extracted markdown. The cache index enables fast queries across fetched pages without repeatedly hitting live websites.
Yes, websh supports optional eager crawling with configurable depth and domain constraints. This allows controlled content discovery across websites and APIs while enforcing asynchronous background extraction.
Extract structured data and view page metadata by navigating to a URL and running stat to see page information. Use find, locate, and cat to discover and capture specific content from the cached markdown.
Yes, you can mount APIs and automate workflows across different web services. websh exposes commands for navigation, search, and data capture, allowing you to interact with APIs as if navigating a Unix-like filesystem.
LvcidPsycheSearch, scrape, and crawl web pages through the Firecrawl API.
Khang5687Synchronize Notion pages to a local cache for offline access and Markdown export.
MaxMiksaNavigate URLs and query web content using Unix-like shell commands.
knoopxFetch web pages, parse HTML with CSS selectors, and query REST APIs.
